In <052f01c56fd8$17d94060$6c62fea9(_at_)ibmrkydk2ufvdd> "John Glube"
<jbglube(_at_)sympatico(_dot_)ca> writes:
[a large snip of Ted's message about closing MARID and the marid
I-Ds deleted]
It is being suggested that sender policy framework was not
before marid.
With respect, the purpose of marid in considering
draft-ietf-marid-protocol-03 and
draft-ietf-marid-mailfrom-00 was to allow for consideration
of that part of the sender policy framework utilizing the
domain in the SMTP mail from as the host to check.
As such, based on the record, in my view the position that
sender policy framework was not before marid is without
merit.
draft-ietf-marid-protocol-03 has a very large number of
incompatibilities with draft-mengwong-spf-*, as far as semantics go.
The draft-ietf-marid-mailfrom-00 I-D was first introduced on Sep 16,
2004. The MARID WG was shut down on Sep 22. No
draft-ietf-marid-helo-00 was ever introduced. At IETF-60, Aug 1-6,
2004, it was decided that the MARID protocol needs to use a new record
version.
So, at no time during MARID was there a MAIL FROM and HELO scope
considered for SPFv1 records. The MAIL FROM scope was introduced only
as MARID was collapsing and never reached consensus.
If you want split hairs, I'll accept that SPFv1 was never adopted by
MARID.
In the alternative, it can be suggested, because these two
drafts did not fully incorporate what is now contained in
the sender policy framework drafts, which endeavours to
document all the myriad versions and usages, both as to
publishing records and host checks, it is therefore
appropriate for the IESG to consider
draft-schlitt-spf-classic-02.
If by "the myriad versions", you mean the versions of the SPFv1 draft
specifications, yes. If you mean to include any of the MARID
protocol, then no.
That position seems to have been accepted by the IESG to
the extent that on 05/25/05 it declared
draft-lentczner-spf-00 dead.
Remember that draft-lentczner-spf-00 was not intended to be used as a
basis for the MARID protocol to use. It was intended to document
SPF-classic. (It failed badly in that regard, and that failure
contributed to the turmoil in the SPF community.)
Presently, the IESG has before it for evaluation purposes
the following:
[spf-classic and the MARID protocol I-Ds deleted]
Now that the drafts have been "polished," the names of the
directorate constituted by the Area Directors to carry out
the focused technical review have been posted.
I think that if you actually check the MARID protocol I-Ds, you will
find that very little polishing has been done on them since the MARID
WG. I would argue that they weren't really polished during that time
either.
To my mind, there are two questions:
* Is the statement in the schlitt draft concerning the
recommended usage of SPFv1 records with any other protocol
appropriate?
I think so.
The IESG in closing marid stated:
The working group chairs and Area Advisor intend to ask
that the editors of existing working group drafts put
forward their documents as non-working group submissions
for Experimental RFC status.
The existing working group drafts did not contain such a
recommendation, but rather proposed spf2.0 and defined
scopes for SMTP mail from and Purported Responsible
Authority, so avoiding the problem.
Right, the MARID protocol avoided the conflict with SPFv1 records, but
also created a new record that superficially looked the same as SPFv1
records but had different semantics. The creation of a new protocol
that is confusingly similar to an existing protocol is bad
engineering.
In draft-lyon-senderid-core-01 and
draft-lyon-senderid-pra-01 the editors have put forward
drafts which allow for backward compatibility with SPFv1
records.
No, those I-Ds put forward incompatible use of SPFv1 records.
However, presumably the editors of
draft-schlitt-spf-classic-02 are stating that based on the
evidence they have gleaned from the wide spread deployment
experience of SPFv1 to date, this would cause unacceptable
false positive levels and as such is not appropriate.
If credible and substantial evidence exists to support this
contention, then it should be presented to the IESG to
support the recommendation.
Some of this evidence has been documented in the spf-classic I-D
itself and more was presented to the IESG. Much more evidence was
discussed on the spf-discuss and MARID lists dating to well before
IETF-60. This would be really hard to miss.
On the other hand, if the concern is theoretical and no
evidence exists of there being any risk of false positives,
Well, if the idea that the world is round is theoretical and no
evidence exists that ...
I'm presuming you are making a joke, and it did make me smile.
Thanks.
* What steps can the SPF council take to facilitate the
IESG in granting the present request to consider
draft-schlitt-spf-classic for standard track?
In my view, in deciding whether to accede to the SPF
council's present request, the IESG will want to consider
the comments by:
* The technical directorate after conducting their focused
technical review of draft-schlitt-spf-classic-02, subject
to any response by the editors of the draft; and,
So far, I have zero evidence that the DEA directorate has conducted
any technical review of the spf-classic I-D, and I have tried very
hard to find such evidence. Please provide any evidence that supports
your claim.
* MAAWG based on draft-newton-maawg-spf-considerations-00,
or any revisions of that draft, as they can provide real
world input as to deployment experience.
Again, I think you need to re-read that I-D because I don't think it
doesn't present any real-word deployment experience.
Should the IESG decide draft-schlitt-spf-classic-02 (or any
revision, along with any lessons learned document) is
worthy of standards track, then from my perspective, given
the decision to close marid, the next step would be to
reconstitute marid and provide
draft-schlitt-spf-classic-02 (?), any lessons learned
document, along with the technical directorate report and
any reports from MAAWG as input documents, so allowing the
working group to quickly proceed with its work in moving
forward with a standard track proposal.
You are certainly free to give the IESG your recommendations. Based
on their rejection of the requests for BoFs for MASS and CLEAR, I very
much doubt that they would try to restart MARID.
It is possible the IESG could decide to skip this step and
simply proceed forward by allowing the draft (with the
related documents) to proceed forward on the standard
track, subject to an Internet wide call for comment.
The IESG regularly accepts individual submissions for standard track.
There certainly wouldn't be anything unusual about taking this route.
Do I think the IESG will accede to the Council's request?
As matters stand today, no.
As previously stated, I think the chances that the IESG will accept
spf-classic-02 for standard track is low. However, I think the
request has increased the chances that it will be accepted in the near
future.
Rather than speculate as to what might be the response of
the IESG, along with the editors of draft-lyon-senderid-core
and draft-lyon-senderid-pra, my suggestion would be to do
two things:
* Ask for input from the technical directorate;
Been there. Done that. Was told no.
* Sit down with Andy Newton and ascertain what
recommendations MAAWG has based on their field testing.
Before I bother Andy about such things, I would need to see evidence
that they have done field testing.
P.S. In an earlier post, Wayne asked if I had any contact
with Meng. I don't know this for certain, but given the
apparent conflict between the schlitt and lyon drafts and
as Meng is listed as an editor on both, I would suggest
that he has decided, as his work is done, the better course
is to simply stand clear and let the chips fall where they
may.
Meng has not only stood clear of the various I-Ds, but he has stood
clear of updating the SPF website (no significant updates since last
fall), posting to spf-discuss (one post in ~6 months), attending SPF
council meetings (missed at least the last 4), communication on IRC
(almost none for a long time), etc. Meng has even stood clear of his
promise to talk with me last week and to attend SPF council meetings.
Yes, Meng appears to be letting the chips fall where they may.
I would much rather see Meng re-engage, as do everyone else I have
talked with. However Meng appears to have abandoned the SPF projected
and all attempts to get him back have failed.
-wayne